Impact of provider type and number of providers on surveillance testing among survivors of head and neck cancers.

BACKGROUND Guidelines for follow-up after head and neck cancer (HNC) treatment recommend frequent clinical examinations and surveillance testing. Here, the authors describe real-world follow-up care for HNC survivors and variations in surveillance testing. METHODS Using Surveillance, Epidemiology, and End Results (SEER)-Medicare data, this study examined a population-based cohort of HNC survivors between 2001 and 2011 Usage of cross-sectional head and neck imaging (CHNI), chest imaging (CI), positron emission tomography (PET), fiberoptic nasopharyngolaryngoscopy (FNPL), and, in irradiated patients, thyroid function testing (TFT) was captured over 2 consecutive surveillance years. Multivariate modeling with logistic regression analyses was used to assess variations by clinical factors, nonclinical factors, number and types of providers seen and their evolution over time. RESULTS Among 13,836 HNC survivors, the majority saw a medical, radiation, or surgical oncologist and a primary care provider (PCP; 81.7%) in their first year of surveillance. However, only 58.1% underwent either PET or CHNI, 47.8% underwent CHNI, 64.1% underwent CI, 32.5% underwent PET scans, 55.0% underwent FNPL, and 55.9% underwent TFT. In multivariate analyses, patients who followed up with more providers and those who followed up with both a PCP and an oncologist were more likely to undergo surveillance testing (P < .007). However, adjusting for providers seen did not explain the variations in surveillance testing rates based on age, race, education, income level, and place of residence. Over time, there was a gradual increase in the use of PET scans and TFT during surveillance years. CONCLUSIONS In this large SEER-Medicare data study, only half of HNC survivors received the recommended testing, and greater compliance was seen in those who followed up with both an oncologist and a PCP. More attention is needed to minimize variations in surveillance testing across sociodemographic groups.